Scalable Backend Engineering • RAG Systems • LLM Applications • System Design • Production Deployment

I am Himanshu Raj, a Computer Science undergraduate at IIIT Nagpur with a specialization in Data Science.

I am focused on building intelligent, scalable, and production-oriented software systems. My work combines strong problem-solving fundamentals with practical engineering across AI systems, backend architecture, system design, cloud deployment, and full-stack development.

I have solved 370+ problems on LeetCode with a rating of around 1700, which has helped me build a strong foundation in data structures, algorithms, and analytical problem solving.

I design and deploy intelligent systems using Retrieval-Augmented Generation (RAG), LLM orchestration, vector search, and containerized deployment. My work focuses on production-ready AI pipelines involving embedding ingestion, semantic retrieval, model evaluation, API integration, and scalable backend design.

• Multimodal RAG architectures
• LangChain-based LLM orchestration
• Vector databases, embeddings, and semantic search
• Gemini API and OpenAI API integration
• Object-Oriented Programming and problem solving
• REST API development using FastAPI, Pydantic, Node.js, and Express
• Docker-based application deployment
• Git, GitHub, and version control workflows
• AWS fundamentals including EC2, S3, and IAM
• Scalable backend systems and distributed application design

• 900+ open-source contributions across GitHub
• 30+ public projects on GitHub
• Contributed to well-known open-source repositories including Axios, Fastify, Transformers, Appsmith, and others
• Internship experience in building scalable systems, production-ready AI pipelines, third-party API integrations, and containerized applications
• Strong competitive programming and DSA foundation with 370+ LeetCode problems solved

Problem: Context-aware AI retrieval across diverse document formats. Built: End-to-end RAG system with ingestion, embeddings, vector search, and response orchestration. Tech: Python, LangChain, FAISS, LLM APIs, Docker

Highlights:

• Processed 1,000+ documents for semantic retrieval
• Reduced retrieval latency by ~30% via optimized vector indexing
• Modular ingestion architecture supporting extensibility
• Containerized deployment for reproducibility

Repository: https://github.com/himaenshuu/Multi_modal_rag-application

Problem: Secure and scalable document management system. Built: Full-stack storage platform with authentication and structured APIs. Tech: Next.js, Node.js, MongoDB, TypeScript

Highlights:

• Implemented JWT-based authentication system
• Handled 10,000+ file metadata records in MongoDB
• Reduced API response time by ~25% through query optimization
• Structured production-ready project architecture

Repository: https://github.com/himaenshuu/smartVault

Problem: Lack of contextual product assistance before purchasing. Built: AI-powered product Q&A and recommendation assistant integrating external APIs. Tech: TypeScript, Next.js, External APIs

Highlights:

• Integrated 500+ product records via API ingestion
• Designed structured recommendation logic pipeline
• Improved response clarity using contextual filtering
• Implemented automated email-based purchase workflow

Repository: https://github.com/himaenshuu/ShopSense

Problem: Limited accessible analytics tools for social media insights. Built: Full-stack analytics dashboard generating structured insights. Tech: TypeScript, Next.js

Highlights:

• Analyzed 5,000+ profile engagement data points
• Built reusable analytics visualization components
• Optimized state management for faster UI rendering
• Designed scalable dashboard architecture

Repository: https://github.com/himaenshuu/InstaLens

Problem: Raw datasets require transformation before analytics usage. Built: Python-based ETL pipeline for cleaning and loading structured data into MongoDB. Tech: Python, MongoDB

Highlights:

• Processed 50,000+ raw records
• Improved data consistency by implementing schema validation
• Reduced manual preprocessing time by ~40%
• Designed reusable transformation modules

Repository: https://github.com/himaenshuu/etl_mongo_pipeline

Problem: Unstructured ecommerce API responses are difficult to use directly for analytics or modeling. Built: Data ingestion and transformation workflow converting API responses into structured datasets. Tech: Python, Jupyter Notebook

Highlights:

• Product feature extraction
• Data normalization and cleaning
• Analytics-ready structured output
• Foundation for recommendation systems

Repository: https://github.com/himaenshuu/Amazon-Product-Intelligence-Data-Builder

Programming Languages: C++ • Python • JavaScript • TypeScript • SQL AI & LLM Systems: LangChain • RAG • Vector Search • Embeddings • Gemini API • OpenAI API • LLM Orchestration Backend Engineering: FastAPI • Pydantic • Node.js • Express.js • REST APIs • API Design Frontend Development: Next.js • React • TypeScript Databases: MongoDB • MySQL • Vector Databases Cloud & DevOps: Docker • AWS EC2 • AWS S3 • AWS IAM • Git • GitHub • Postman

• Build systems that are clear, scalable, and maintainable
• Prefer modular architecture over tightly coupled code
• Design APIs and services with production use cases in mind
• Use containerized environments for reproducibility and deployment
• Optimize only where performance is measurable
• Write code that is understandable for both humans and machines
• Treat documentation as part of engineering, not an afterthought

Strong engineering is not only about making things work. It is about making systems reliable, understandable, scalable, and useful in the real world.

I am interested in opportunities across:

• Software Engineering
• Backend Engineering
• Full-Stack Development
• AI Engineering
• LLM-Based Product Development
• RAG and Intelligent System Architecture

I am particularly interested in building products at the intersection of AI system architecture, scalable backend engineering, and real-world software deployment.

• Strong problem-solving mindset
• Curiosity and fast learning ability
• Adaptability across technologies and domains
• Team management and leadership skills
• Clear communication in English and Hindi
• Product-oriented thinking with engineering depth